Multimodal biometric authentication based on score level fusion of finger biometrics

Feature fusion-based multimodal biometrics has become an increasing interest to many researchers in recent years, particularly for finger biometrics. There are, however, many challenges in fusing multiple feature sets, as the case with Canonical Correlation Analysis (CCA) and Multi-set Canonical Correlation Analysis (MCCA). How to extend them to fuse multiple feature sets is a significant problem in general. In this paper, we propose a novel multimodal finger biometric method, which provides feature fusion approach called linear discriminant multi-set canonical correlation analysis (LDMCCA). It combines finger vein, fingerprint, finger shape and finger knuckle print features of a single human finger. Compared with CCA and MCCA, LDMCCA contains the class information of the training samples and represents the fused features more efficiently and discriminatively in few dimensions. The experimental results on a merged multimodal finger biometric database show that LDMCCA is beneficial to fuse multiple features as well as achieves lower error rates than the existing approaches.

With the increasing demand for information security and security regulations all over the world, biometric recognition technology has been widely used in our everyday life. In this regard, multimodal biometrics technology has gained interest and became popular due to its ability to overcome a number of significant limitations of unimodal biometric systems. In this paper, a new multimodal biometric human identification system is proposed, which is based on a deep learning algorithm for recognizing humans using biometric modalities of iris, face, and finger vein. The structure of the system is based on convolutional neural networks (CNNs) which extract features and classify images by softmax classifier. To develop the system, three CNN models were combined; one for iris, one for face, and one for finger vein. In order to build the CNN model, the famous pertained model VGG-16 was used, the Adam optimization method was applied and categorical cross-entropy was used as a loss function. Some techniques to avoid overfitting were applied, such as image augmentation and dropout techniques. For fusing the CNN models, different fusion approaches were employed to explore the influence of fusion approaches on recognition performance, therefore, feature and score level fusion approaches were applied. The performance of the proposed system was empirically evaluated by conducting several experiments on the SDUMLA-HMT dataset, which is a multimodal biometrics dataset. The obtained results demonstrated that using three biometric traits in biometric identification systems obtained better results than using two or one biometric traits. The results also showed that our approach comfortably outperformed other state-of-the-art methods by achieving an accuracy of 99.39%, with a feature level fusion approach and an accuracy of 100% with different methods of score level fusion.

Feature fusion-based multimodal biometrics has become an increasing interest to many researchers in recent years, particularly for finger biometrics. In this paper, a novel multimodal finger biometric method based on Multi-set Canonical Correlation Analysis (MCCA) is proposed. It combines finger vein, fingerprint, finger shape and finger knuckle print features of a single human finger. The proposed approach transforms multiple unimodal feature vectors into sets of canonical correlation variables, which represent fused features more efficiently in few dimensions. The experimental results on a merged multimodal finger biometric database show that the proposed approach has significant improvements over the existing approaches. It is beneficial to fuse multiple features as well as achieves lower error rates.KeywordsMultimodalFingerFeature fusionMulti-set Canonical Correlation Analysis

Biometric is used to automate the measurement of biological data. The measurement and recording of the physical characteristics of an individual for the use in subsequent personal identification. Multimodal biometric authentication is system which uses two or more biometric trait. It provides more immunity against spoofing. To reduce feature vector size and to improve genuine acceptance rate proposed technique is developed. In this paper fusion of Left Fingerprint, Right Fingerprint, Iris, Palmprint are experimented. Multimodal biometrics having different level fusion such as score level fusion, feature level fusion, decision level fusion. In this paper score level fusion is considered for experimentation. Score level fusion contains more gratified and worthful information. Here different score proportions are experimented and performance efficiency is measured using genuine acceptance ratio (GAR). True acceptance rate of recognition is increased because of multiple biometrics characters. In proposed technique features are extracted using Thepade's sorted ternary block truncation coding. Using TSTBTC and matching score proportion Iris: Palm print: Left Fingerprint: Right Fingerprint(40∶2∶1∶1) gives better performance as indicated by higher GAR values observed by 71.86%.

The biometric based personal verification system is technique used to calculating physical or behavioral characteristics of human. The biometric system is fundamental alternative of ID cards, passwords, passports, driving licenses. Biometric schemes have some restrictions in the terms of accuracy, acceptability, distinctiveness, universality. The approaches for combining two or more biometric have attracted increasing attention of researchers. Aim of combining two or more biometric is to increasing the accuracy of system. The combination of two or more biometric scheme is known as “biometric fusion”. The biometric fusion is classified four types those are 1) sensor level fusion, 2) feature level fusion, 3) score level fusion and 4) decision level fusion. In this paper we analyze the performance of fusion at the different level in multimodal biometric.

Single sensor-based multi-modal biometrics is a promising approach that offers simple system construction, low cost, and wide applicability to real situations such as CCTV footage-based criminal investigations. In multi-modal biometrics, fusion at the score-level is a popular and promising approach, and data qualities that affect the matching score of each modality are often incorporated as a quality-dependent score-level fusion framework. This paper presents a very large-scale single sensor-based multi-quality multi-modal biometric score database called MultiQ Score Database version 2 to advance the research into evaluation, comparison, and benchmarking of score-level fusion approaches using both quality-independent and quality-dependent protocols. We extracted gait, head, and height modalities from the OU-ISIR Gait Database and introduce spatial resolution (SR), temporal resolution (TR) and view as quality measures that significantly affect biometric system performance. We considered seven and 10 scaling factors for SR and TR, respectively, with four view variations. We then constructed a database comprising approximately 4 million genuine and 7.5 billion imposter score databases. To evaluate this database, we set two different protocols, and provided a set recognition accuracy for state-of-the-art approaches using protocols for both quality-independent and quality-dependent schemes. This database and the evaluation results will be beneficial for score-level fusion research. Additionally, we provide detailed analysis of the recognition accuracies associated with gait, head, and height modalities in different spatial/temporal resolutions and views. These analyses may be useful in criminal investigation research.

Score level fusion of multimodal biometrics using triangular norms

Advancements in multimodal biometrics, which integrate multiple biometric traits, promise to enhance the accuracy and robustness of identification systems. This study focuses on improving multimodal biometric identification by using fingerprint and finger vein images as the primary traits. We utilized the "NUPT-FPV" dataset, which contains a substantial number of finger vein and fingerprint images, which significantly aided our research. Convolutional neural networks (CNNs), renowned for their efficacy in computer vision tasks, are used in our model to extract distinct discriminative features. Specifically, we incorporate three popular CNN architectures: ResNet, VGGNet, and DenseNet. We explore three fusion strategies used in security applications: early fusion, late fusion, and score-level fusion. Early fusion integrates raw images at the input layer of a single CNN, combining information at the initial stages. Late fusion, in contrast, merges features after individual learning from each CNN model. Score-level fusion employs weighted aggregation to combine scores from each modality, leveraging the complementary information they provide. We also use contrast limited adaptive histogram equalization (CLAHE) to enhance fingerprint contrast and vein pattern features, improving feature visibility and extraction. Our evaluation metrics include accuracy, equal error rate (EER), and ROC curves. The fusion of CNN architectures and enhancement methods shows promising performance in identifying multimodal biometrics, aiming to increase identification accuracy. The proposed model offers a reliable authentication system using multiple biometrics to verify identity.

In the biometrics, the technologies grow day by day and the security also increased related to that technologies. The fingerprint was the most intensively researched in the field of biometrics system due to permanence and uniqueness features which made varies of different peoples. The paper addressing many stages, in addition to the primary stages of any biometrics system the fusion of unimodal system was used in order to improve the performance of the system. The double enhancement techniques were used to make the images very clear by Histogram Equalization and Fast Fourier Transformation (FFT). The feature extraction was conducted using three techniques which called Zernike Moment (ZM), Hu-Moments (Hu) and Gray-Level Co-occurrence Matrix (GLCM) that categorized to statistical and texture features. The matching between these features was performed using the Euclidean distance to find the scores matrix. Additionally, the fusion as the most modern technique was used to improve the performance of the biometrics system which performed in this work by feature level and score level fusions. The feature level fusion by using concatenation and score level fusion by using Weight sum rule strategy led to improve the performance of the system. The system was evaluated by False Accepted Rate (FAR), False Rejected Rate (FRR), Equal Error Rate (EER) and Genuine Accept Rate (GAR). The results show that, the fusion gave the most efficiency results compared with individual system. The work was tested on four datasets such as Fingerprint Verification Competition (FVC2000), (FVC2002), (FVC2004) and our department datasets which called KVK dataset. The best results were achieved by FVC2002 with maximum GAR reached to 98.45% and minimum EER of 1.54% as compared with other datasets and existing works.

We constructed a large-scale multi-quality multi-modal biometric score database to advance studies on quality-dependent score-level fusion. In particular, we focused on single sensor-based multi-modal biometrics because of their advantages of simple system construction, low cost, and wide availability in real situations such as CCTV footage-based criminal investigation, unlike conventional individual sensor-based multi-modal biometrics that require multiple sensors. As for the modalities of multiple biometrics, we extracted gait, head, and the height biometrics from a single walking image sequence, and considered spatial resolution (SR) and temporal resolution (TR) as quality measures that simultaneously affect the scores of individual modalities. We then computed biometric scores of 1912 subjects under a total of 130 combinations of the quality measures, i.e., 13 SRs and 10 TRs, and constructed a very large-scale biometric score database composed of 1,814,488 genuine scores and 3,467,486,568 imposter scores. We finally provide performance evaluation results both for quality-independent and quality-dependent score-level fusion approaches using two protocols that will be beneficial to the score-level fusion research community.

A multimodal biometric system amalgamates the information from multiple biometric sources to alleviate the limitations in performance of each individual biometric system. In this paper a multimodal biometric system employing hand based biometrics (i.e. palmprint, hand veins, and hand geometry) is developed. A general combination approach is proposed for the score level fusion which combines the matching scores from these hand based modalities using t-norms due to Hamacher, Yager, Weber, Schweizer and Sklar. This study aims at exploring the potential usefulness of t-norms for multimodal biometrics. These norms deal with the real challenge of uncertainty and imperfection pervading the different sources of knowledge (scores from different modalities). We construct the membership functions of fuzzy sets formed from the genuine and imposter scores of each of the modalities considered. The fused genuine score and imposter scores are obtained by integrating the fuzzified genuine scores and imposter scores respectively from each of the modalities. These norms are relatively very simple to apply unlike the other methods (example SVM, decision trees, discriminant analysis) as no training or any learning is required here. The proposed approach renders very good performance as it is quite computationally fast and outperforms the score level fusion using the conventional rules (min, max, sum, median) The experimental evaluation on a database of 100 users confirms the effectiveness of score level fusion. The preliminary results are encouraging in terms of decision accuracy and computing efficiency.

Recently multimodal biometrics technology that employs more than two types of biometrics data has been popularly used for person authentication and verification. In particular, the score-level fusion approach which combines matching scores from unimodal systems to make final decision has gained lots of attentions. In most of these works, however, they assume all the matching scores to be of the same quality. This assumption may cause the problem not to reflect such situation that the qualities of the matching scores from certain unimodal systems are relatively low. To deal with this problem, we propose the RBF based score-level fusion approach which incorporates the quality information of the scores in developing classification models. According to our experimental results, the proposed method using quality information showed its superiority in the performance of person authentication to the usual RBF based score-level fusion without using quality information.

In this paper, we make a first attempt to combine face and iris biometrics using an efficient local appearance feature extraction method based on steerable pyramid (S-P), to captures the intrinsic geometrical structures of face and iris image, it decomposes the face and iris image into a set of directional sub-bands with texture details captured in different orientations at various scales. Local information is extracted from S-P sub-bands using block-based statistics to reduce the required amount of data to be stored. The obtained local features are combined at the score level for developing a multimode biometric approach, which is able to diminish the drawback of single biometric approach as well as to improve the performance of authentication system. We combine a face database FERET and iris database CASIA (version 1) to construct a multimodal biometric experimental database with which we validate the proposed approach and evaluate the multimodal biometrics performance. The experimental results reveal the multimodal biometric authentication is much more reliable and precise than single biometric approach.

Multimodal biometric system fuses information from multiple biometric modalities to overcome limitations of unimodal biometric system. This fusion significantly enhances the performance of the system. One of the ways of fusing information for multimodal biometrics is score level fusion. In this paper, a novel score level fusion method is proposed. Here, fusion at score level is formulated as an optimization problem. The paper proposes a genetic algorithm (GA) based approach to solve this optimization problem. It minimizes the distances between an aggregated score list and each input score list from individual biometric modality. The proposed GA based method uses weighted Spearman footrule distance metric to compute the distance between a pair of score lists. Superiority of the proposed method over several state-of-the-art score level and rank level fusion methods is demonstrated experimentally.

Biometrics technology stands as one of the major backbones that had united biosciences and technology representing an instrument for security and forensics researchers to develop more accurate, robust and confident systems. Starting from uni-modal biometrics as finger print, face, speech and iris passing through multimodal biometrics based on uni-biometrics fused by different fusion techniques as feature level, score level and decision level fusion techniques, biometrics were still one of the most investigated technologies. From here in this paper, we tried to build the base for researchers whom are interested in biometric systems through introducing a comparative study of most used and known uni- and multimodal biometrics such as face, iris, finger vein, face and iris multimodal, face, finger print and finger vein multimodal. Through this comparative study, a comparative model is based on principal component analysis feature extractor and Euclidean distance matcher applied using MATLAB. This model was trained and tested in two different modes homogenous data using SDUMLA-HMT database and heterogeneous mode extracting 106 frontal single face image from CASIA-FACEV5 while the reminder biometrics under consideration from SDUMLA-HMT. Feature level and score level fusions were tested in both modes on all multimodal systems under consideration.